Attributing blame: how political parties in Germany leverage cooperative federalism

AbstractThis study investigates how political parties used the federal structure of government for discursive blame attribution strategies in parliamentary debates during the Covid-19 crisis. The analysis focuses on the German case which is considered an embodiment of cooperative federalism. Largely intertwined responsibilities and joint decision making provide incentives for self-serving blame attribution strategies. The empirical investigation includes a qualitative content analysis of 212 parliamentary debates in the Bundestag and the 16 state parliaments. Overall, 2067 statements were manually coded and integrated into a novel dataset. The data reveal a more diverse discursive toolkit of blame attribution strategies than commonly conceptualised. The study demonstrates that parties, especially when they are involved in intergovernmental bodies and coalition governments, resort to ‘softer’ forms of blaming. The vertical integration of the party system also creates an effective blame barrier, containing self-serving strategies even during the prolonged crisis and several election campaigns.Keywords: Cooperative federalismblame attributionparty competitionparliamentary debatesCovid-19 AcknowledgementsWe thank the two reviewers for their helpful and constructive comments on earlier versions of this article and Akseli Paillette-Liettilä, Jonathan Röders, Polina Khubbeeva, Yannis Wittig, and Marek Wessels for their valuable research assistance.Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 The dataset, codebook, and further project documentation are available at the data repositorium of GESIS: https://doi.org/10.7802/2627.2 The existing literature uses both the term ‘blame attribution’ and the term ‘responsibility attribution’. While the literature on voters and their perceptions of multilevel systems generally uses the term ‘responsibility’, the more institutionally oriented contributions focus on the term ‘blame’. However, federalism research usually means by the term ‘responsibility’ the legal or constitutional competences of federal entities in federal systems. Therefore, we prefer to use the term ‘blame’ to delineate formal responsibilities from discursive ascriptions.3 Infection Protection Act of 20 July 2000 (BGBl. I: 1045), last amended by Article 8b of the Act of 20 December 2022 (BGBl. I: 2793).4 It was not possible to analyse all parliamentary debates on Covid-19. Based on the protocols of the plenary sessions in the Bundestag and the 16 Landtage, we initially marked all procedures related to managing Covid-19 that were debated there between 1 February 2020 and the federal elections on 26 September 2021. In total, we have identified 3117 procedures in this period.5 Regular discussions as well as the joint specification of definitions and coding instructions ensured a common understanding among the team members on how to code the debates. The Landtage were coded by five coders. Each coder reviewed his or her coded statements after some time and suggested changes to the original coding. This affected about 6% of the cases. These changes were discussed by the team and implemented accordingly. To ensure inter-coder reliability, every tenth debate (total: 21 debates) was again coded by a second coder. The second coder agreed with 78% of the originally coded strategies. Finally, each statement was checked by (at least) a second coder: 8% of the cases were adapted or deleted. The 10 debates in the Bundestag were coded by two coders. Initially, they coded five debates each. They then swapped debates and coded again the other five debates. The results were compared, differences were discussed, and the coding finalised. A third team member eventually checked all coded text sections: 6% of the cases were re-coded or deleted.6 In Germany, parliamentary party groups determine the deputies who take the floor for them in the debates (Müller et al. Citation2021: 381). Therefore, we can assign the deputies, speaking in the debates, to their respective party. Moreover, we treat the government representatives in the debates as speakers for their parties. This has been decided in the awareness that due to the different roles, the statements of the deputies can be more competitive and polarised than the ones of government representatives.7 Since the CDU and CSU do not compete at either level of government, and it is not clear whether the speakers in the Bundestag represent only the CDU, CSU, or both due to their joint parliamentary party group, we have decided to consider them together here. Empirically, the CSU is also not an outlier vis-a-vis comparable party branches of the CDU. The CDU’s Länder branches with governing responsibility exhibit 53% for ‘passing responsibility’ and 47% for ‘finding a scapegoat’. Including the CSU, the numbers slightly change to 52% and 48%.8 To determine the exact time periods of the three Covid-19 waves, we refer to the official information of the Robert Koch-Institute (RKI) (Tolksdorf et al. Citation2021). The RKI is the federal government’s institute for surveillance and prevention of infectious diseases and pandemic preparedness. The first wave lasted from 02 March 2020 to 17 May 2020, the second wave from 28 September 2020 to 28 January 2021, and the third wave from 01 February 2021 to 13 June 2021.9 When parties refer to a specific Land or Länder, we had to decide how to determine their party affiliation because they are usually governed by a coalition. We use the party affiliation of the respective minister president as distinction. As they represented the Länder in the meetings of the MPK with the chancellor, the minister presidents played the key role in federal crisis management and especially in the media reporting about it. Therefore, citizens primarily associated the Länder with their minister presidents.10 The five Länder were Baden-Württemberg, Rhineland-Palatine, Saxony-Anhalt, Mecklenburg-Western Pomerania, and Berlin. The elections in Thuringia were postponed to 2024 as the necessary two-thirds majority for the dissolution of the state parliament (Art. 50, 2; Constitution of Thuringia) could not be achieved.Additional informationFundingThis work was supported by Volkswagen Foundation as part of the project ‘Political cohesion under conditions of fiscal scarcity – German federalism in the time of COVID-19 ‘[project number 2021000121] and by the Berlin University Alliance as part of the project ‘The Laws of Social Cohesion (LSC)’.Notes on contributorsAntonios SourisAntonios Souris is Postdoctoral Researcher at the Chair of German Politics at Freie Universität Berlin. In his dissertation, he investigated EU policy coordination in Germany’s federal system. His research focuses on comparative federalism, parliaments, and the policy areas of transport and housing. [antonios.souris@fu-berlin.de]Sabine KroppSabine Kropp is Professor of German Politics at Freie Universität Berlin. Her primary field of research is comparative federalism and multilevel politics, parliamentarism and public administration, with an emphasis on Germany and post-Soviet countries. Her recent book Emerging Federal Structures in the Post-Cold War Era (co-edited with Soeren Keil, Palgrave Macmillan 2022) investigates emerging and regressing federal structures in unconsolidated federal systems. [sabine.kropp@fu-berlin.de]Christoph NguyenChristoph Nguyen is Lecturer at the Chair of German Politics at Freie Universität Berlin. He received his PhD from Northwestern University. His research focuses on the intersection between affect, ideas, and politics and the use of quantitative, qualitative, and mixed-methods in political science. [christoph.nguyen@fu-berlin.de]